Fig. 2

Receiver operating characteristic (ROC) curve analyses assessing the ability of preoperatively measured markers to diagnose 30-day mortality. Receiver operating characteristic curve (ROC) analyses for the prediction of 30-day death. AUC confidence interval and significance tests using Bootstrap resampling (times = 500). The blue shading denotes the bootstrap estimated 95% confidence interval with the AUC. Model 1 comprises history of cerebrovascular disease, CREA, time of operation based on differences observed between groups on recruitment (Table 1). ROC area (AUC): 0.708 (95%CI, 0.546-0.836). Model 2 comprises the MST1, history of cerebrovascular disease, CREA, time of operation. ROC area (AUC): 0.823 (95%CI, 0.700-0.912). P values (P = 0.02) relate to the comparison of Model 1 to the Model 2 using DeLong test for matched ROC curves.